An improved and robust DNA immunization method to develop antibodies against extra-cellular loops of multi-transmembrane proteins

Abstract

Multi-transmembrane proteins are especially difficult targets for antibody generation largely due to the challenge of producing a protein that maintains its native conformation in the absence of a stabilizing membrane. Here, we describe an immunization strategy that successfully resulted in the identification of monoclonal antibodies that bind specifically to extracellular epitopes of a 12 transmembrane protein, multi-drug resistant protein 4 (MRP4). These monoclonal antibodies were developed following hydrodynamic tail vein immunization with a cytomegalovirus (CMV) promoter-based plasmid expressing MRP4 cDNA and were characterized by flow cytometry. As expected, the use of the immune modulators fetal liver tyrosine kinase 3 ligand (Flt3L) and granulocyte-macrophage colony-stimulating factor positively enhanced the immune response against MRP4. Imaging studies using CMV-based plasmids expressing luciferase showed that the in vivo half-life of the target antigen was less than 48 h using CMV-based plasmids, thus necessitating frequent boosting with DNA to achieve an adequate immune response. We also describe a comparison of plasmids, which contained MRP4 cDNA with either the CMV or CAG promoters, used for immunizations. The observed luciferase activity in this comparison demonstrated that the CAG promoter-containing plasmid pCAGGS induced prolonged constitutive expression of MRP4 and an increased anti-MRP4 specific immune response even when the plasmid was injected less frequently. The method described here is one that can be broadly applicable as a general immunization strategy to develop antibodies against multi-transmembrane proteins, as well as target antigens that are difficult to express or purify in native and functionally active conformation.

Introduction

The development of potentially therapeutic monoclonal antibodies against extracellular loops or domains of multi-transmembrane proteins continues to be technically challenging. Oftentimes the extracellular regions of these proteins are highly structured, poorly immunogenic, small, or well-conserved. Solubilized protein or synthetic peptides as antigens frequently fail to recapitulate native epitopes, and therefore when used as immunogens result in antibodies that are incapable of binding to the native conformation of the protein.^1,2 However, immunization with whole cells expressing the target of interest has been successful in generating antibodies to 7‑transmembrane G-protein coupled receptors (GPCRs) and 12-transmembrane transporters.^1,3 While intact cells and membrane preparations can be used as immunogens,^4-6 their success is oftentimes hindered by low target antigen expression, non-native epitope structure, and protein instability. DNA immunization has a long history of having been successfully used to develop monoclonal antibodies (mAbs) and is particularly amenable to the in vivo expression of structurally native full-length proteins in the membrane-bound state,^7-9 which makes this approach an appealing one for the development of potentially therapeutic mAbs for the treatment of unmet medical needs, including cancer, type II diabetes and autoimmune disorders.^10,11

The immune response following DNA Immunization is in part dependent on the level of target antigen that is expressed, which can vary as a result of the efficiency of DNA uptake into the cells and the promoter used. The human cytomegalovirus immediate-early (CMV-IE) promoter is capable of driving high-level transcription in a wide variety of mammalian cell types.¹² Other groups have shown that the CMV-IE enhancer with a chicken β-actin promoter, modified to include the splice acceptor sequence of the rabbit β-globin gene (CAG) promoter, is capable of inducing an increased and sustained level of protein expression in comparison to other promoters, including CMV,^13,14 as notably observed in mice.¹⁵ This has been attributed to the strong activity of the chicken β actin promoter, as well as more efficient processing of the transcript resulting from the inclusion of the rabbit β globin splice acceptor site.^14,16 The pCAGGS plasmid, which includes the CAG promoter, woodchuck post-transcription regulatory element (WPRE) and the rabbit β-globin polyadenylation site, together can yield higher protein expression vs. the CAG promoter alone.¹⁷

Multidrug resistance protein 4 (MRP4), a 12-transmembrane multidrug transporter, was selected as a test antigen for developing a DNA based immunization strategy to readily identify antibodies against multi-transmembrane antigens. MRP4 is a 1325-amino acid protein that is structurally similar to other ABC transporter family members, having two membrane-spanning domains, each with six transmembrane helices.¹⁸ While MRP4 is highly conserved across many species, it does not share significant sequence similarity with other ABC transporters.¹⁹ Functionally, MRP4 is instrumental in the cellular transport of a wide range of molecules, including many anti-cancer and anti-viral compounds, and has broad substrate specificity.²⁰

Here, we describe the identification and characterization of four unique monoclonal antibodies that bind specifically to extracellular epitopes of MRP4. Multiple attempts to develop specific cell surface binding anti-MRP4 antibodies using either phage libraries or conventional hybridoma technology were unsuccessful. These efforts included the use of detergent-solubilized membrane protein or MRP4 expressed on the surface of cells, and displayed on baculovirus, as antigens. While conventional approaches failed, hydrodynamic tail vein (HTV) injection of DNA encoding the MRP4 gene, in conjunction with immune modulators, proved to be a successful strategy to identify antibodies with the desired binding properties: specific binding to native conformational epitopes expressed on the surface of MRP4-expressing cells. Furthermore, use of the pCAGGS plasmid DNA as a technical enhancement increased the overall efficiency of this approach as it allowed us to immunize less frequently without compromising the overall immune response to the target antigen.

Results

MRP4 target validation and reagents for antibody development

MRP4 is a 12-transmembrane ABC transporter with six extracellular loops, including two major loops consisting of 19 and 33 amino acids, respectively (Fig. 1A), as predicted by the DAS (Transmembrane Prediction Server) program. Full-length MRP4 cDNA was sub-cloned with a c-terminal fusion to green fluorescent protein (GFP) under control of the CMV promoter and a neomycin selectable marker. Stable 293 cells expressing MRP4-GFP (293-MRP4GFP) were generated. The androgen-sensitive human prostate adenocarcinoma cell line LnCAP was used as a model for endogenous expression of the MRP4 protein. LnCAP cells expressing a dox-inducible shRNA targeting the MRP4 transcript were also developed. Relative levels of total and cell surface MRP4 on these cell lines were evaluated by immunofluorescence and western blot, and found to be downregulated upon treatment with doxycycline (Fig. 1B and C). Cells expressing MRP4 exhibited resistance to micromolar levels of the cytotoxic agent irinotecan (CPT-11), a topoisomerase-1 inhibitor commonly used as a chemotherapeutic drug. Control 293 cells endogenously expressing minimal levels of MRP4, and LnCAP cells in which MRP4 was downregulated by shRNA induction, exhibited decreased viability in response to increasing drug levels (Fig. 1D). Together, these results validated the utility of these cell lines as tools for the screening and identification of antibodies against MRP4.

Figure 1. Functional characterization of MRP4 overexpressing and knock-down cell lines. (A) Schematic representation of MRP4, a 12-transmembrane ABC transporter. (B) Western blot analysis of MRP4 knock-down in 293 and LnCAP cells were performed using an anti-MRP4 antibody as described in methods. Cells were treated with and without doxycyclin (50 ng/mL, 48 h). (C) Analysis of cell surface expression of MRP4 protein in 293 and parental LnCAP cell lines. After treatment with doxycyclin (24 h), cell surface expression was downregulated. (D) 293, 293-MRP4GFP (high MRP4 levels with MFI of 80–100) and LnCAP cells were then exposed to irinotecan (CPT-11) at different concentrations in culture medium for 48 h. After 48 h of drug exposure, cells were assessed for viability using CellTiter-Glo and IC₅₀ values were determined.

Strategies used to develop anti-MRP4 antibodies

Various approaches were taken to identify monoclonal antibodies capable of binding to extracellular epitopes of MRP4 (Table 1). Pre-incubation of phage libraries with 293-shMRP4 KO cells, followed by panning on 293-MRP4GFP cells, failed to yield antibodies that were positive by FACS. Balb/c and MRP4 knockout mice, for conventional hybridoma development, were hyper-immunized with 293-MRP4GFP cells as an initial strategy. MRP4-specific FACS binding was not observed with the immune sera and therefore fusions were not performed. Sera from Balb/c mice immunized with LnCAP cells, and MRP4 KO mice immunized with purified MRP4 protein expressed in insect cells, also failed to show MRP4-specific FACS binding. Selection of MRP4-specific Fab fragments by phage display against MRP4 displayed on baculovirus was done as previously described.²¹ Eight phage-derived antibodies bound MRP4 solubilized in dodecylmaltoside by ELISA, and three were found to specifically immunoprecipitate MRP4, but none of these showed binding to cells overexpressing or endogenously expressing MRP4 by FACS. Of the various strategies taken, only DNA immunizations resulted in antibodies that bound specifically to cell-surface expressed MRP4 (Table 1).

Table 1. Methods used for the development of anti-MRP4 monoclonal antibodies.

Immunogen	Species	Route	Dose per animal or panning round	Result
293-MRP4 cells	Balb/c, C57Bl/6-MRP4 KO mice, Phage	IP	1x107 cells	Non-specific FACS binding, No FACS+ ELISA +ve antibodies
LnCAP cells	Balb/c mice	IP	5x106 cells	Non-specific FACS binding
Purified MRP4 protein from insect cell lysates	C57Bl/6-MRP4 KO mice	Footpad	2 µg	No FACS+ve, Western blot +ve antibodies
Baculovirus-displayed MRP4 protein	Phage library screening	None	None	No reactivity, No FACS+ve, only ELISA +ve antibodies
pRK-MRP4myc DNA	Balb/c mice	HTV	50 µg	1 mAb FACS +ve
pRK-MRP4myc DNA + mFlt3L DNA, mGMCSF DNA	Balb/c, C57Bl/6-MRP4 KO mice	HTV	50 µg	3 mAb (Balb/c) FACS +ve

IP, intraperitoneal; HTV, hydrodynamic tail vein.

Comparison of DNA immunization with and without the use of immune modulators

HTV-based DNA immunization was performed by injecting 50 µg of pRK-MRP4myc plasmid DNA (Fig. 2A) diluted in lactated Ringer’s solution in a volume equivalent to 10% of the body weight of the mouse over a period of 4–8 s (Fig. 2B). This bolus injection resulted in forced extravasation of liquid volume, including DNA, which was then taken up into major organs adjacent to the vena cava, primarily the liver. DNA immunizations were performed weekly with 50 µg of plasmid DNA. Immune sera from 20 Balb/c mice, evaluated five days after the fourth DNA injection, showed minimal to no difference in binding to 293-MRP4GFP cells and untransfected 293 cells (Fig. 2C). New immunizations of Balb/c and C57BL/6 MRP4 knockout mice were performed by HTV injection as previously described. These mice received a priming dose of 10 µg pORF-mFlt3L DNA (HTV) one week prior to immunization with the pRK-MRP4myc DNA. In addition, these mice received weekly injections of both 2.5 µg of pORF-mGM-CSF DNA and 50 µg of pRK-MRP4myc DNA (co-immunized). Sera from 16 of the 20 mice showed a differential FACS shift between 293-MRP4GFP cells and untransfected 293 cells following the mFlt3L priming and four co-injections of MRP4 and mGM-CSF DNA (Fig. 2D). These results together demonstrated that the addition of Flt3L and GM-CSF as immune modulators significantly improved not only the overall immune response in these mice, but importantly the induction of antibodies capable of recognizing native extracellular epitopes.

Figure 2. DNA immunization using the hydrodynamic tail vein injection technique and characterization of immune responses. (A) Plasmid maps of pRK-MRP4myc, pCAGGS-DsRed-Luciferase (CAGGS-Luc) and pCMV-Luciferase-IRES-MRP4GFP (CMV-Luc-MRP4GFP). (B) Plasmid DNA diluted in a volume of lactated Ringer’s solution equivalent to 10% of the mouse’s body weight was injected through the tail vein of the mouse over 4–8 s. The large volume travels through the vena cava through the hepatic portal vein and the liver temporarily expands to accommodate the volume, forcing uptake of the DNA into the hepatic cells. Balb/c (#726–745, #861–870) or C57BL/6 MRP4 knockout mice (#851–860) were immunized weekly with 50 µg pRK-MRP4myc DNA without (C) or with (D) the use of immune modulators. Mice #851–870 were injected with 10 µg pORF-mFlt3L DNA as a priming dose, then injected weekly with both 50 µg pRK-MRP4myc DNA and 2.5 µg pORF-mGM-CSF DNA. Serum samples were evaluated five days following the fourth dose of MRP4 DNA, and screened on 293 cells transfected with MRP4 DNA (blue) and 293 control cells (green).

Characterization of anti-MRP4 antibodies

Splenocytes from mice showing MRP4-specific antibody titers by FACS were used for fusion and hybridoma development. Four monoclonal antibodies, designated 2E4, 4D11, 13H2, and 14E7, respectively, were identified from three independent fusions. These antibodies demonstrated the ability to bind native MRP4 expressed on the surface of transfected 293 cells (Fig. 3A) and MRP4 endogenously expressed on LnCAP cells (Fig. 3B). A FACS-based competitive binding assay using Alexa-488 labeled 13H2 antibody for detection revealed that only unlabeled 13H2 antibody could compete for MRP4 binding on 293-MRP4-expressing cells (Fig. 3C). Interestingly, experiments using labeled 2E4, 4D11 and 14E7 in FACS competition with the four unlabeled antibodies showed similar results in that each antibody did not compete out signal from the others (Table 2), suggesting that the four anti-MRP4 antibodies were directed against different extracellular epitopes. To assess function blocking activity of anti-MRP4 antibodies, we incubated 293-MRP4GFP cells with varying concentrations of CPT-11 in the presence and absence of anti-MRP4 antibody. As seen from Figure 3D, all four MRP4 antibodies did not have any influence on cell viability in response to increasing drug levels. These results suggest that MRP4 antibodies examined in this study do not interfere with the transport function of MRP4 upon their binding to cell surface epitopes of MRP4.

Figure 3. Characterization of anti-MRP4 monoclonal antibodies. Anti-MRP4 monoclonal antibodies (1 µg/mL) specifically bound to 293 cells stably overexpressing MRP4GFP (A). Binding of 2E4 antibody on methanol-fixed 293-MRP4GFP stable cells by fluorescence microscopy. FACS analysis was performed on endogenously MRP4-expressing LnCAP cells (B). Competitive binding assay was done by directly conjugating Alexa-647 fluorophore to 13H2 and adding excess of MRP4 antibody clones to compete for the 13H2 epitope (C). MRP4 antibodies were characterized for their ability to inhibit transport of MRP4 substrate (CPT-11) (D). Cell viability of 293-MRP4GFP (moderate MRP4 levels with MFI of 40–50) was measured by adding varying concentration of CPT-11 inhibitor in the absence or presence (20 μg/mL) of anti-MRP4 antibody in culture medium for 72 h and then cells were assessed for viability as described in Figure 1.

Table 2. MRP4 antibodies bind non-overlapping epitopes of MRP4 as analyzed by competitive binding assay using Alexa-647 labeled antibodies.

		Alexa-647 labeled antibodies (0.5 μg/mL)
		2E4	13H2	4D11
Excess unlabeled Ab (20 μg/mL)	2E4	4.5	14.7	21.6
	13H2	26.8	6.3	25
	4D11	23.8	15.2	4.6
	14E7	26.3	17.1	21.9

Decreasing the frequency of immunizations increases process efficiency

To achieve robust FACS positive titers using the method described above, weekly HTV injections beyond 5–7 weeks is typically required. This extended immunization regime is both labor intensive for our technical staff and physiologically stressful for the mice. Protein expression resulting from a CMV-driven promoter plasmid was shown to be stable for 1–2 d in various in vitro cell lines.²² In contrast, CAG promoter-driven expression plasmids, such as pCAGGS, were shown to result in a sustainable response beyond several days.²³ We therefore sought to determine whether this prolonged expression could be achieved in vivo and potentially streamline our HTV immunization process. As nude mice are the standard strain used for bioluminescence imaging, we injected these mice with plasmids encoding the luciferase gene under the control of either the CMV or CAG promoter using HTV injection. Luciferase activity in mice injected with the CMV promoter plasmid was undetectable within a week, whereas activity was sustained beyond 9 weeks with the CAG promoter plasmid (Fig. 4).

Figure 4. Comparison of in vivo luciferase expression following hydrodynamic tail vein injection of plasmid DNA containing different promoters. Nude mice were injected with 50 µg of pRK-Luciferase (CMV) or pCAGGS-DsRed-Luciferase (CAGGS), then injected with 200 µL of 25 mg/mL D-luciferin (i.p.) and imaged using a CCD camera at 24 h and weekly following HTV injection.

We then generated a CMV-based MRP4-IRES-luciferase construct and quantitated the protein levels in vivo by bioluminescence imaging. Nude mice received a single HTV injection of 50 µg of either the luciferase-encoding pCAGGS plasmid or the CMV promoter plasmid encoding both MRP4 and luciferase with an IRES element (Luciferase-IRES-MRP4) (Fig. 2A). Luciferase expression was monitored by bioluminescence 1 d, 2 d, and 7 d following the DNA injections (Fig. 5A). Photon counts per min per cm² of the area around the liver confirm that expression was 2 to 3 fold higher with the pCAGGS plasmid one day following injection as compared with the CMV promoter plasmid. In addition, the expression of luciferase was undetectable by day 2 with CMV plasmid (Fig. 5B).

Figure 5. Luciferase expression following hydrodynamic tail vein injection of plasmid DNA using CMV and CAG promoters. Nude mice were injected with 50 µg of CMV-Luc-MRP4GFP or CAGGS-Luc, followed by injection of 200 µL of 25 mg/mL D-luciferin (i.p.) and imaged using a CCD camera at 24 h, 48 h, and 1 week following HTV injection (A). Photon counts per min per cm² of observational area (red circles) were calculated and compared using M3 Vision software (B).

To assess reproducibility of these results, the experiment was repeated in Balb/c mice. Luciferase expression was again monitored 1 d, 2 d, and 7 d post-injection (Fig. 6A). Ex vivo imaging two days post-injection showed that luciferase expression was observed only in the liver (Fig. 6B). Photon counts of the area surrounding the liver, measured 1–2 d post-injection, showed a greater than 99% decrease in luciferase expression (Fig. 6C). Collectively, these results suggested that immunization with CMV promoter DNA, while capable of inducing protein expression in Balb/c mice, is short-lived and would require repeated and frequent injections to maintain detectable expression of a target protein.

Figure 6. In vivo luciferase expression analysis in Balb/c mice. Five Balb/c mice (#230–234) were injected with 50 µg of pCMV-Luciferase-IRES-MRP4GFP, then injected with 200 µl of 25 mg/mL D-luciferin (i.p.) and imaged using a CCD camera at 24 h, 48 h, and 1 week following HTV injection (A). Mouse #232 was euthanized and organs were surgically removed 48h following HTV injection, then irrigated with 50 µL D-luciferin and imaged using a CCD camera (B). Photon counts per min per cm² of observational area (red circles) were calculated and compared using M3 Vision software (C).

The above results suggested that prolonged in vivo expression of a target antigen could be achieved with the pCAGGS plasmid. Given this potential advantage of the pCAGGS plasmid vs. CMV, we then compared both plasmids in parallel for their ability to effectively induce specific antibody titers. Balb/c mice were immunized with plasmids encoding MRP4, with either CMV or CAG promoters, on a weekly, bi-weekly, or monthly basis. Following eight weeks of injections with the CMV-promoter plasmid, FACS-positive titers on LnCAP cells were observed only in the group receiving the bi-weekly injections. In contrast, all groups of the pCAGGS-immunized mice showed a robust immune response in comparison to their CMV-immunized cohorts (Fig. 7). Together, these results strongly suggest that CAGGS promoter-driven expression would require less frequent dosing compared with CMV promoter-driven plasmids, therefore resulting in a significant technical resource savings and overall representing a more efficient DNA-based immunization strategy for therapeutic antibody discovery.

Figure 7. FACS analysis of mice immunized with MRP4 expressing plasmids consisting of either the CMV or CAG promoters. Bleeds from mice immunized with the CMV plasmid (weekly, bi-weekly and monthly) were analyzed on LnCAP cells for binding to MRP4 expressed on the cell surface. Specific binding was not observed with the immune and naive mouse sera (A). Bleeds from mice immunized weekly, bi-weekly and monthly with the pCAGGS DNA showed robust anti-MRP4 titers. Differential binding was observed with all pCAGGS immunized mice as compared with naïve controls (B).

Discussion

Multi-transmembrane proteins are targets of interest for the development of potentially therapeutic monoclonal antibodies, but often prove to be challenging or impossible to express in a native state. While immunization with soluble/solubilized protein, intact cells, membrane lysates or linear/cyclized peptides is possible, these approaches are not always successful in producing antibodies that are capable of binding to native extracellular epitopes or present additional challenges in the screening process. On the contrary, DNA-based immunization strategies have been shown by numerous groups to provoke a robust immune response against native conformational epitopes of proteins for both antibody development^24,25 and protective immunity.^26,27 Intramuscular injection of DNA, followed by site-specific in vivo electroporation, has also been described as an effective method to produce monoclonal antibodies.²⁸ While this method has been previously shown elsewhere to be successful in generating a specific immune response, we have observed otherwise, e.g., consistently lower levels of protein expression with this method, or the need to extend the immunization timeline to achieve an optimal immune response (Vij R and Hongo J, et al., unpublished results). The HTV injection method has been used to express high levels of protein in the liver for vaccination and mouse model development,²⁹ and to induce protective immunity. As it has been previously shown that HTV DNA delivery can effectively result in a high titer and specific antibody response, we used this technique to develop antibodies against the multi-transmembrane ABC transporter, MRP4. Use of HTV DNA immunization succeeded in generating MRP4-specific FACS positive binding antibodies where multiple alternative immunization strategies failed (Table 1). Success with HTV delivery was also observed with a structurally similar and highly conserved 12-transmembrane glucose transporter (Tan C and Hongo J, et al., unpublished results). The inclusion of immune modulators prior to and during immunization dramatically improved the overall immune response (Fig. 2C and D). Administration of mFlt3L-encoding plasmid DNA one week prior to the target DNA injection, following by co-immunization with independent plasmid DNAs encoding mGM-CSF and MRP4, resulted in a notable increase in the immune response, likely due in part to the expansion and recruitment of antigen-presenting dendritic cells and a subsequent enhancement of the T cell response, consistent with results observed in previously described DNA vaccine studies.^30,31 We succeeded in identifying only one antibody from the cohort of mice immunized with plasmid DNA alone (without mFlt3L priming or mGM-CSF boosting), and only 1 of the 20 mice immunized in this group showed an immune response. In contrast, the addition of mFlt3L and mGM-CSF in the immunization strategy resulted in the majority of mice showing good immune responses. Fusion of splenocytes pooled from multiple animals in this group provided an increased immune repertoire from which to draw, ultimately leading to the identification of three additional MRP4 specific antibodies (Fig. 3A–C). Hence, where other methods failed, HTV DNA immunization with plasmids encoding MRP4, mFlt3L and mGM-CSF as immune modulators, proved to be a successful strategy to identify a panel of monoclonal antibodies capable of binding to MRP4 expressed on the surface of cells.

DNA immunization with CMV promoter-driven plasmids often results in expression of target antigen with a half-life of 24–48 h (Fig. 4),²² and therefore a technical consideration is that more frequent and labor-intensive immunization strategies would be necessary to overcome this short-lived expression. The pCAGGS plasmid, contains both a CAG promoter (shown to yield higher expression, especially in mouse cells^15,23) and WPRE, which helps to stabilize and transport mRNA to the cytoplasm to assist in translation.¹⁷ It could therefore be hypothesized that the presence of both elements in the plasmid encoding the target protein would result in increased target antigen expression, and subsequently, an increased immune response. Our results comparing the CMV promoter and pCAGGS plasmids for HTV DNA immunization support this hypothesis. A direct promoter comparison measured by bioluminescence imaging showed an impressive difference in the levels of luciferase expression in nude mice over time: the duration of CMV promoter plasmid-driven protein expression was 24–48 h, while CAG promoter-driven protein expression was observed beyond nine weeks (Fig. Four and 5A). Furthermore, luciferase expression of the pCAGGS plasmid showed ~2–3 fold higher peak levels compared with the CMV promoter-based plasmid. This observation was consistent in both nude and Balb/c mice (Figs. 5B and 6C). Therefore, use of the pCAGGS plasmid would provide a significant technical resource advantage in that immunizations could be performed less frequently without compromising the overall antibody response (Fig. 7).

Our data demonstrate that HTV DNA immunization with the pCAGGS plasmid has the potential to provoke a strong immune response to a variety of challenging multi-transmembrane proteins while dosing as infrequently as once a month. We demonstrated that this method can be applied successfully, using MRP4 as an example, to develop potentially therapeutic monoclonal antibodies to a structurally complex target antigen for which stable, properly folded native protein was difficult to obtain, and where cell-based immunizations and other antibody discovery strategies have failed.

Materials and Methods

Animals

Mice used in these studies were maintained in an AAALAC accredited animal facility. All experiments were performed in compliance with Genentech’s Institutional Animal Care and Use Committee and OLAW Guidelines.

Plasmids

MRP4 gene was cloned into the pRK vector using EcoR1 and Sal1 sites with either a GFP or His tag. The coding region of human MRP4 was subcloned into the pRK (Genentech) expression vector with either a myc or GFP fusion at the c-terminus of the MRP4 protein. Commercially-obtained human prostate mRNA (Clontech 637215) was used to generate the desired PCR product. The PCR conditions were as follows: 94°C for 4 min and 40 cycles of 94°C for 30 s, 52°C for 30 s and 72°C for 5 min. The pCAGGS-DsRed-Luciferase plasmid was generated by cloning DsRed-Express (Clontech 632412), 2A,³² and a synthesized DNA fragment containing firefly luciferase (GeneBank accession number AY603757.1) into a pCAGGS plasmid described previously.³³ The pCAGGS-MRP4 construct was amplified by PCR using 5′-GCGGCCGCGCCACCATGCTGCCCGTGTACCAGG-3′ and 5′-CTCGAGTCACAGTGCTGTCTCGAAAATAGTTAAGG-3′. Both the pCAGGS vector³³ and PCR product were digested with Not1 and Xho1 and ligated. All final constructs were confirmed by sequencing.

Cell lines

MRP4GFP stable cell lines (293-MRP4GFP) were initiated by transient transfection of the pRK-MRP4GFP plasmid into 293 cells cultured in a 60-mm tissue culture dish at 50% confluence. Transient transfections were performed using the FuGENE 6 transfection reagent (Roche Diagnostics 11814443001). Transfected cells were grown in culture medium for 3 d, then trypsinized and split into 15 60-mm tissue culture dishes containing DMEM supplemented with 1.4–1.8 mg/mL G418 (Sigma-Aldrich G8168). Resulting colonies were trypsinized and isolated using sterile glass cloning rings (Bellco Glass) and then expanded into tissue culture flasks containing selection medium. Potential MRP4GFP expressing colonies were screened by fluorescence microscopy, and selected colonies were expanded. Western blots were performed using α-MRP4 antibody (GeneTex, Clone M4I-80; Irvine, CA) on the selected colonies to confirm fusion protein expression. Stable cell lines were then maintained in medium supplemented with 0.4 mg/mL G418. Two different stable 293 cells expressing MRP4-GFP (293-MRP4GFP) were generated and one at moderate level (mean fluorescence intensity, MFI of 40–50 based on GFP intensity) and another one at high level expression of MRP4 (MFI of 80–100)

CellTiter-Glo assays

Effects of CPT-11 were measured on 293-MRP4GFP and LnCAP cells and IC₅₀ determinations were made using the CellTiter-Glo assay (Promega Corp G7570). LnCAP cells (1x10⁴) were plated into 96-well plates and allowed to adhere overnight at 37°C. CPT-11 was diluted to varying concentrations in culture media, added to the cells in 96-well plates and incubated for 48 h. After incubation, cell viability was measured by aspirating the media and adding CellTiter-Glo reagent. Plates were mixed, incubated at 37°C for 30 min and luminescence measured after using the Envision plate reader (Perkin Elmer).

Functional blocking assays were performed by plating 293-MRP4GFP with moderate expression of MRP4 (1 × 10⁴) in 96-well plate and grown overnight at 37°C. Media was aspirated and 100 μL of varying concentration of CPT-11 was added along with 20 μg/mL of antibody to each well. Plates were incubated for 72 h and luminescence was measured using Envision plate reader as described above.

DNA Immunizations

Balb/c mice (Charles River Laboratories) or C57BL/6/MRP4.ko mice (Genentech) were immunized with 50 µg of pRK-MRP4myc plasmid DNA with or without pORF-mFlt3L and pORF-mGM-CSF (in-licensed from St. Jude Children’s Research Hospital, Memphis TN, through MTA),³⁴ diluted in lactated Ringer's solution via HTV injection as previously described.³⁵ The mice receiving the immune modulators were injected with 10 µg pORF-mFlt3L DNA as a priming dose, followed by weekly injections of both 50 µg pRK-MRP4 DNA and 2.5 µg pORF-mGM-CSF DNA. For the promoter comparison studies, Balb/c mice (Charles River Laboratories) were immunized with 10 µg pORF-mFlt3L, followed by weekly, bi-weekly or monthly injections of 50 µg of either pCMV-Luciferase-IRES-MRP4GFP or pCAGGS-MRP4 plasmid DNA plus 2.5 µg pORF-mGM-CSF diluted in lactated Ringer's solution via HTV injection.

Hybridoma development

Spleens were harvested three days following the last HTV immunization. Splenocytes from these mice were fused with X63-Ag8.653 mouse myeloma cells (American Type Culture Collection PTA-8315) by electrofusion (Cytopulse or BTX, Harvard Apparatus, Holliston, MA) and incubated at 37°C in the presence of 5–7% CO₂ overnight in Dulbecco's Modified Eagle's Medium (DMEM; Lonza 12–709F) supplemented with 10% fetal bovine serum (FBS; Hyclone SH30070.02), 4.5 g/L glucose, 25 mM HEPES, 0.15 mg/mL oxaloacetic acid (Sigma O9504), 100 µg/mL pyruvic acid (Sigma P3662), 0.2 U/mL insulin (Sigma I6634), 2 mM L-glutamine (Invitrogen 25030164), 100 U/mL penicillin, 100 µg/mL streptomycin (Penicillin-Streptomycin; Invitrogen 15140122), NCTC-109 (Lonza 12-923E), NEAA (Invitrogen 11140076), before addition to 96-well plates supplemented with 5.7 µM azaserine and 100 µM hypoxanthine (HA; Sigma-Aldrich A9666). Supernatants were screened for IgG production by ELISA and FACS 11 d post-fusion. All IgG-secreting hybridomas demonstrating MRP4-specific binding by FACS were then expanded and subcloned by limiting dilution. Final hybridoma clones demonstrating significant FACS positive shifts following two rounds of subcloning were then expanded for large-scale production in bioreactors (Integra Biosciences CELLine1000). Hybridoma supernatants were then purified by Protein-A affinity chromatography as previously described.³⁶

Bioluminescence imaging

Nude mice were injected with 50 µg of pRK-Luciferase (Genentech) or pCAGGS-DsRed-Luciferase (Genentech). Mice were anesthetized using isoflurane (Butler Schein Animal Health 2091966), injected intraperitoneally (i.p.) with 200 µL of 25 mg/mL D-luciferin (Invitrogen L2912) using a tuberculin syringe, placed in a light tight box and imaged using a CCD camera capturing light in the range of 500–700 nm. Initial characterization studies used a cooled intensified CCD camera from Stanford Photonics with typical image acquisition times of 10 s. All subsequent bioluminescence studies were imaged on the Photon Imager from BioSpace Lab, Paris, France. Mice were imaged at 24 h and weekly following HTV injection. During image acquisition, animals were maintained on anesthesia using a nose cone delivery system and their body temperatures regulated using a thermostatically controlled warm pad. For ex vivo imaging, organs were surgically removed from a pRK-immunized mouse 24 h following HTV injection, rinsed with PBS then irrigated with 50 µL D-luciferin. Photon counts per min per cm² of observational area were calculated and compared using M3 Vision software (BioSpace Lab).

Balb/c mice were injected with 50 µg of pCMV-Luciferase-IRES-MRP4GFP (Genentech) then injected i.p. with 200 µL of 25 mg/mL D-luciferin and imaged using a CCD camera at 24 h, 48 h, and 1 week following HTV injection. Prior to imaging, the abdominal area was depilated with Nair to reduce light scattering from the fur. For ex vivo imaging, one mouse was euthanized and organs were surgically removed 48 h following HTV injection, rinsed in PBS then irrigated with 50 µL D-luciferin and imaged using a CCD camera. Photon counts per min per cm² of observational area were then calculated and compared using M3 Vision software.

Flow cytometry analysis

Sera from Balb/c or C57BL/6 MRP4 knockout mice was taken five days following the fourth injection of MRP4 DNA and screened on both 293 control and 293-MRP4GFP cell lines. Hybridoma supernatants and purified antibodies were screened on 293 control, 293-MRP4GFP and LnCAP cell lines (with and without Doxycycline treatment).

Cells were collected from flasks/dishes, washed with phosphate-buffered saline (PBS), and added to 96-well U-bottom plates (BD Falcon 353077) at 200,000–500,000 cells per well. Sera were diluted 1:100 in PBS containing 1% FBS (Hyclone SH30070.02); hybridoma supernatants were screened neat. Samples were added to cells (100 µL/well) and incubated at 4°C for 30–60 min. Plates were then centrifuged (1200 rpm, 5 min, 4°C) and washed twice with PBS/1% FBS (200 µl per well). Phycoerythrin-conjugated goat anti-mouse IgG Fc (Jackson ImmunoResearch 115–116–071; 100 µL diluted in PBS) was then added and the plates incubated at 4°C (covered) for 30 min. After the final wash, the cells were fixed in PBS containing 1% formalin, and read using a FACSCalibur flow cytometer (BD). Mean fluorescence intensity (MFI) of each sample was then measured using the FlowJo software (Treestar, Inc.).

Western blots and immunofluorescence

Whole cell extracts from 293, 293-MRP4-GFP and LnCAP (± doxycycline treatment) were run on 4–20% gradient precast SDS-PAGE gels (Invitrogen EC6028), and proteins transferred to Immobilon-P membrane by electro-blotting using the iBlot apparatus (Invitrogen). The membranes were incubated in blocking buffer (Pierce 37542) for 30 min, washed three times with Tris-buffered saline (TBS), incubated with an anti-MRP4 antibody (GeneTex, Clone M4I-80) for 3–4 h at room temperature. The membranes were then washed three times with TBS and binding detected with an anti-rat HRP antibody (Pierce 37585).

LnCAP and 293-MRP4 cells were cultured in 6-well glass slides overnight as described above. The medium was then aspirated and cells fixed by the addition of 0.5 mL of cold methanol (15 min, -20°C). Following aspiration of the methanol, the cells were washed three times with PBS. Immunofluorescence was then measured as previously described.³⁷